Air Pollution and Short-Term Mortality in Beijing

Little empirical evidence exists on the health costs of air pollution in China, one of the most polluted countries in the world. Unsurprisingly, the lack of reliable data on pollution levels and health outcomes impede research. Because the pollution-health relationship is likely non-linear, it is difficult to extrapolate from existing high quality studies in developed countries to ascertain health costs. We address this deficiency by obtaining new data on Beijing’s daily mortality April 2008-April 2013 from the Chinese Center for Disease Control and Prevention. We combine these data with daily pollution measures from the US Embassy in Beijing, which records particulate matter of 2.5 microns or less in width (PM 2.5). We find that after controlling for weather conditions, year, month, and day of week fixed effects, daily PM2.5 indeed predicts daily mortality, particularly deaths from cardiovaslular disease. A 100 μg/m3 increase in daily PM2.5 is associated with 7 deaths daily, among them 4 cardiovascular deaths, and 0.8 respiratory deaths. Furthermore, deaths among less-educated and outdoor workers show a stronger relationship to PM2.5 levels. Notably, the relationship is robust to controlling for the official measure of Beijing’s air pollution, the average daily air pollution index (API), despite the fact that PM2.5 is measured by 1 monitor at the US embassy whereas API (and mortality) combine data from across the Beijing metropolitan area. Indeed, Beijing’s API does not have a significant relationship to mortality once AQI at the Embassy is accounted for. Our finding supports previous research arguing for measuring PM 2.5 and reporting it promptly to the public.